Nvidia Corporation

Semiconductor Components Industries, LLC

Yan Zhai

Farzin Aghdasi

Parthasarathy Sriram

Fnu Ratnesh Kumar

Anil Ubale

Subhashree Radhakrishnan

In various examples, sensor data—such as masked sensor data—may be used as input to a machine learning model to determine a confidence for object to person associations. The masked sensor data may focus the machine learning model on particular regions of the image that correspond to persons, objects, or some combination thereof. In some embodiments, coordinates corresponding to persons, objects, or combinations thereof, in addition to area ratios between various regions of the image corresponding to the persons, objects, or combinations thereof, may be used to further aid the machine learning model in focusing on important regions of the image for determining the object to person associations.

Determining associations between objects and persons using machine learning models

In various examples, sensor data&#x2014;such as masked sensor data&#x2014;may be used as input to a machine learning model to determine a confidence for object to person associations. The masked sensor data may focus the machine learning model on particular regions of the image that correspond to persons, objects, or some combination thereof. In some embodiments, coordinates corresponding to persons, objects, or combinations thereof, in addition to area ratios between various regions of the image corresponding to the persons, objects, or combinations thereof, may be used to further aid the machine learning model in focusing on important regions of the image for determining the object to person associations.

A method of performing lane detection without the use of a frame buffer may include capturing image frames with an image sensor in a vehicular imaging system. Feature extraction circuitry in the vehicular imaging system may analyze a frame to detect features corresponding to possible lane markers. The features may be extracted and stored in memory for further processing, while the rest of the frame may not be stored in memory and may not undergo further processing. Processing circuitry may perform a first estimation of lane marker location based on a continuous feature that is present in multiple different image frames and may perform a second estimation of lane marker location based on multiple features in a single image frame that form a connected feature. The processing circuitry may determine the presence of a lane marker based on the continuous feature and the connected feature.

Growing community of inventors

Santa Clara, CA, United States of America

Yan Zhai